Return to BSD News archive
Path: sserve!newshost.anu.edu.au!munnari.oz.au!news.Hawaii.Edu!ames!agate!howland.reston.ans.net!europa.eng.gtefsd.com!usenet From: niemidc@oasis.gtefsd.com (David C. Niemi) Newsgroups: comp.os.386bsd.bugs Subject: Re: Disk geometry x-lation foils multiple BIOS p Date: 17 Jun 1993 17:55:47 GMT Organization: GTE Federal Systems Division Lines: 44 Distribution: world Message-ID: <1vqb73$pt1@europa.eng.gtefsd.com> References: <JTSILLA.93Jun16095528@dec5200b.ccs.northeastern.edu> Reply-To: niemidc@oasis.gtefsd.com NNTP-Posting-Host: hengist.lab.oasis.gtegsc.com In article 93Jun16095528@dec5200b.ccs.northeastern.edu, jtsilla@dec5200b.ccs.northeastern.edu (James Tsillas) writes: [ . . . ] > e) Boot NetBSD again to copy kernel. Things boot but when I > enter 'copy' it complains that it cannot find a label and > that the secondary boot block is bad. I suspect this is the > previosuly reported bug in handling driver who's controllers > do disk geometry translation. > > f) Figuring that NetBSD will have to wait I boot DOS again and > use fdisk to set partition (1) as active. I then install > MS-DOS to partition (1) without any problems. > > g) When I try to boot partition (1) I am somehow finding a boot > program which tries to boot partition (2). Since I couldn't > copy the kernel in step (e) I get repeated 'Could not find > kernel: blah' messages. > > h) Figuring that my primary boot block has been overwritten by > NetBSD braindeath (yes, I was slightly upset at that). I went > to my trusty AMI BIOS and used low level format. You need to use some sort of boot manager to coexist between the two. I got things to work using the OS/2 boot manager from the beta release, but there are many alternatives. I could also use the DR-DOS fdisk, like your f) above, and it worked fine. My guess is that you had something wrong with the disk geometry you supplied. I just gave NetBSD the "fake" geometry consistently, as when I (previously) tried to give it the "real" geometry nothing worked. Also, if you look at the numbers Seagate gives you, the numbers for the "fake" geometry don't add up the the same total as they do for the "real" one, so I wouldn't trust the "real" one anyway. At any rate, NetBSD can definitely grok the "fake" geometries correctly, at least with my hard disk controller (a very nice DTC 2290 EISA). I, too, have an AMI (1990) BIOS, but its "low-level format" does not work! I'm getting a replacement for other (VLB) reasons. --- David C. Niemi: David.Niemi@oasis.gtegsc.com My opinions are those of my fuzz-brained, cat-sniffing Norwegian Elkhound.